Federal Court accepts Wayback Machine as evidence of content on a specific date

A couple decades ago I used Wayback Machine to win a DMCA takedown case that almost bankrupted me.

An American military contractor (I'm Canadian) stole a large chunk of content from my site, search/replaced my company name for theirs. They even made really slick PDF files complete with graphics (that were possibly actually theirs) to make it look legit. When confronted, their legal dept claimed I actually stole it from them. I was FUMING!

The worst thing was that when confronting their lead techie, I asked him to describe a particular bit of terminology that is very specific to me and my business. He claimed he couldn't because it was proprietary. Of course it was because he simply didn't know anything about that particular term. I guess he thought that would drag things out long enough that I'd eventually drop it. I didn't.

I then demonstrated that this bit of terminology is attributed to me in a published IEEE document that also predated the content appearing on his web page. But for whatever reason, their web page and all their marketing collateral said that they created it all.

Needless to say, they ended up removing the content and paying damages. I couldn't afford to take it further, and given their significant wealth, would have easily quashed me if I tried. I won and got paid for it, so I accepted it was probably the most I could possibly get.

I got the idea based on the "poor man's copyright," which is when musicians, for example, postal mail lyrics to themselves and then store the unopened envelope in a safe in order to prove, via postmark, that the copyright date is legit.

I suspect github could also be used this way. Anything that keeps a strongly maintained timeline.

So, in the case of NTP vs. RIM, part of the legal battle was fought in the UK. I was called in as an expert witness on a detail of minutia, as to when certain things had been said publicly, which could be used as evidence of prior art.

They had pulled up copies of the comp.mail.sendmail FAQ, for which I was the maintainer in 95-96. They asked me if a certain section of the document they found was a clear and accurate representation of the way the document actually existed at the time. They contacted me and not Eric Allman, because his role in the maintenance of the FAQ didn’t cover the period of time they were interested in, even though I believe he actually wrote that section years earlier. What was important to them was the state of the document at the time, not who wrote what when.

I want back to various archive sources, and confirmed that what they had appeared to me to be an accurate copy of the file from that point in time.

I was literally one day away from having to appear in UK court through a videoconference facility at the offices of a local law firm here in Austin, TX, where I was actually going to testify to that statement. But they settled the case instead.

Ultimately, I think the critical part of the process is to get an expert in the field to be willing to testify as to the material you are showing them, and not so much the source of that material.

In order to have the systems testify for themselves, they have to use forensic methods throughout. As much as I love them, that’s not the Wayback Machine.

Failing that, I don’t see any alternative to having an expert human who can testify that the material presented to them by the Wayback Machine is actually an authentic representation of the content from the time.

I can see where it could be argued that content in the Wayback Machine could be edited before being presented as evidence; I know the current case specifically excludes the question of whether the content in the scrapes was edited or not bu what would be a viable scheme for verifying that the content in a scrape hasn't been edited?